Causes of temporal variability of lead in domestic plumbing systems.

Sources of lead in drinking water are primarily lead pipe, lead/tin solder, and brass fixture materials.Lead levels in the water depend upon many solubility factors, such as pH, concentrations of substances such as inorganic carbonate, orthophosphate, chlorine, and silicate, the temperature, the nature of the pipe surface, etc. Physical factors, time, and chemical mass transfer are significant in governing lead levels in nonequilibrium systems. The diameter and length of lead pipe is extremely important, as well as the age and chemical history of the solder and brass fixtures. Analytical variability is not particularly significant relative to between-site and within-site variability. Knowledge of temporal variability at each site is necessary to define a statistically valid monitoring program. An analysis of published data covering repetitive measurements at a given site show that the variability of lead concentration at each site tends to be characterized by the frequent occurrence of 'spikes'. Variability expressed as approximate relative standard deviations tends to be of about 50 to 75% in untreated water, regardless of the mean lead concentration. The distributions are frequently nonnormal for small numbers of samples. Monitoring programs must incorporate controls for the causes of the within-site and between-site variability into their sampling design. The determination of necessary sampling frequency, sample number, and sample volume must be made with consideration of the system variability, or the results will be unrepresentative and irreproducible.